A novel website fingerprinting method for malicious websites detection

Abstract

To detect malicious websites which usually together with other normal websites on the cloud. Thus, precise detection method is needed. In this work, we investigate the website fingerprinting (WFP) methods and propose a novel WFP model to classify malicious websites on the cloud. The proposed model can recognize websites from traffic collected with multi-tab and achieve excellent results with almost 200 positive classes. Furthermore, the method displays a promising result in detecting online gambling websites.